In the new Atlantic Equities report, analyst Simon Clinch points out that 60% of Upstart’s revenue comes from Cross River Bank and labels this a concentration risk. I have heard this line before in discussions about Upstart, including on the Motley Fool. But is it really a risk?
The Business Model
As far as I understand it, Upstart’s marketing and partnership with Credit Karma drives traffic to their website, Upstart.com. Once they are on the website, a potential borrower can apply for a loan, and if they are approved, Upstart relies on Cross River Bank (or Finwise Bank) to originate it. Cross River Bank then most likely sells the loan back to Upstart, who in turn sells it to investors (banks and institutions). And why wouldn’t an investor buy a slice of these loans? The KBRA data shows us that these loans have low delinquency rates and target “hidden prime” borrowers traditionally neglected by FICO.
In David Girouard’s interview with the Motley Fool, he called Cross River Bank and Finwise Bank ‘conduits’ to the investors who will ultimately buy the loans. I think this is apt language, and importantly, I think these conduits could be replaced if there is an issue and they no longer want to do / can do business with Upstart for whatever reason.
This business model is in conjunction with Upstart’s growing ‘white labelling’ revenue. Here, banks use the Upstart AI to decide whether they want to approve loans or not. In this situation, the bank can also sell the loan back to Upstart once they originate it, which then again is sold to investors by Upstart. Based on the figures I’ve got, about 75% banks go this route (down from 83%), but maybe even more will hold on to the loans once they figure out that the Upstart algorithm is very good at keeping delinquencies low.
Asking the Right Questions
So instead of saying things like “Cross River Bank is a customer risk”, we should be asking better questions:
- Will capital markets keep on buying Upstart loans?
- Related: Will Upstart’s AI tech keep delinquencies low in the case of a bad macroeconomic event?
- Will Upstart’s AI tech be as effective in other flavors of credit, such as auto refinance, mortgages, and others?
- Will lending volume decrease?
Right now, I am satisfied with the answers to those questions. Upstart has a huge TAM in unsecured personal loans alone, so I’m not worried about loan volume, and I would imagine investors are very happy with current delinquency rates. But as a responsible investor, I’m going to keep an eye on the next few earnings reports.
I don’t have a background in finance (as a matter of fact, I’m a schoolteacher) but I figured this was a worthy discussion, and I think I have my facts right. Please respond if I’m missing something about CRB being a concentration risk or any other of my points. Thanks for the wonderful board, it’s changed my life.
PeterHay